Base pointer to derived object?

Question

If Child inherits from parent, and I do:

Child* c = new Child();
Parent* p = c;

what is the purpose of p ? If Child contained additional data members, do they follow on from p , just that p cannot access them?

I am trying to understand the difference between whats actually in memory and what the pointer can access, depending on its static type.

Answer 1

In the tiny snippet of code shown, it's not possible to tell what MAY be the purpose of p . However, in general, using pointer or reference to a base class to store objects that are polymorphic in nature is a common practice. For example, the classic example of shapes:

// Base class for all shapes. 
class Shape
{
  ... 
};

class Circle: public Shape
{
  ...
};

class Square: public Shape
{
  ...
};

Now, if we want to store a bunch of shapes, we can use:

std::vector<Shape*> shapes;

shapes.push_back(new Circle);
shapes.push_back(new Square);

and then do:

for(auto &s : shapes)
{
   s.Draw();    // Draw shapes on screen... 
}

The vector needs to hold pointers or references, so that the type is determined by the actual object stored. If we made a vector std::vector<Shape> shapes , then we'd get "slicing", where any extra information stored in the Circle (such as radius ) or Square (such as side ) would be lost.